Flusso is a visual workflow automation platform. It lets you build, run, and monitor automated processes that connect AI models, APIs, file storage, messaging services, and your own business logic — all without writing code.
Flusso is a self-hosted automation platform built for teams that need precise control over their data and AI workflows. You design workflows as sequences of connected nodes, where each node performs a specific action: calling an AI model, fetching data from a knowledge base, sending an email, making an HTTP request, and much more.
Workflows can be triggered manually, on a schedule, by an incoming webhook, or by events emitted from other workflows. Results are monitored in real time, and every execution step is logged for inspection.
Build workflows in two editor modes:
The Agent node lets you integrate large language models (OpenAI, Anthropic Claude, local Ollama models, and more) directly into your workflows. Agents can search the web, query your knowledge bases, call external tools via the Model Context Protocol (MCP), and maintain memory across conversations.
Upload documents (PDF, text) to a Knowledge Base. Flusso indexes them using vector embeddings, allowing the RAG (Retrieval-Augmented Generation) node to perform semantic search and return the most relevant excerpts to feed into your AI agents.
Every workflow starts with a trigger. Flusso supports manual triggers, cron-based schedules, incoming webhooks (with optional authentication), and event-based triggers that listen for events emitted by other workflows.
The Monitor and Timelines pages give you a real-time view of running and recently completed workflows. Every step logs its input, output, duration, and any errors — making it easy to diagnose failures.
The Approval and Manual Input nodes pause a workflow and wait for a human response before continuing. Pending tasks are surfaced on the Tasks page for easy action.
After logging in, the interface is divided into two areas: the left sidebar for navigation and the main content area for each section.
| Section | Description |
|---|---|
| Dashboard | Overview stats, recent errors, and activity charts for all your workflows. |
| Workflows | Create, edit, and manage all your automated workflows. |
| Knowledge Bases | Upload and manage document collections for AI-powered search. |
| Tasks | Respond to pending approvals and manual input requests from running workflows. |
| Monitor | Live view of all workflow runs across your account. |
| Timelines | Gantt-style timeline showing step execution durations per workflow. |
| Settings | Configure AI providers, email, integrations, and memory backends. |
Follow these steps to create and run your first workflow:
Throughout Flusso, you can reference the output of any previous node using the double-brace syntax. This allows you to pass data from one step to the next dynamically.
The step_key is the unique identifier assigned to each node in your workflow.
Keys are displayed on every step card and can be customised in the step configuration drawer.
You can click on data pills in the output panel to automatically insert the correct syntax into
any configuration field.
| Status | Meaning |
|---|---|
| Active | The workflow is live and will execute when its trigger fires. |
| Draft | The workflow is being edited and will not respond to triggers. |
| Archived | The workflow is retained for reference but cannot be triggered. |
| Status | Meaning |
|---|---|
| Completed | The workflow finished successfully. |
| Failed | An error occurred during execution. Check the run detail for the error message. |
| Running | The workflow is currently executing. |
| Waiting | The workflow is paused, awaiting human approval or input. |
| Queued | The workflow is waiting for a worker to pick it up. |